Total Synthesis of (+)-Compactin (ML-236B) - American Chemical

methy1encdioxy)methyllycaconitine. Total Synthesis of (+)-Compactin (ML-236B). Nai-Yi Wang, Chi-Tung Hsu, and Charles J. Sih*. School of Pharmacy ...
0 downloads 0 Views 238KB Size
J. Am. Chem. SOC.1981, 103, 6538-6539

6538

AH --#3

6H

$

Lycoctonine

The apparent efficacy of 1 as a therapeutic agent for the treatment of hypercholesterolemia in humans and the enhanced activity of mevinolin, 2, in in vitro animal models precipitated our interest to devise general synthetic methodologies for the preparation of this important family of substituted hexahydronaphthalene lactones. Herein, we report the first synthesis of (+)-compactin. Central to our synthetic plan is the development of a viable route to 3. This key intermediate not only possesses strategically situated oxygen functionalities for the eventual elaboration of the trunsoid conjugated diene but also allows the construction of the highly sensitive 8-hydroxy lactone of 1 during the final stages of the synthetic sequence. We projected that 3 might be prepared by the combination of conjugate addition of the functionalized mixed cuprate 43,4to the enone 55 and trapping of the resulting kinetic enolate with a suitable electrophile. In turn, 5 may be derived from the readily available trans-dione (6).6

O C H3

8

7 Brwniine

Gipctonine

dimethylated product of gigactonine and that given for delphatine, prepared from lycoctonine, are so similar that the minor differences in their chemical shifts could result from solvent effects. On the basis of our results, it seems likely that the dimethlated product prepared from gigactonine and the sample of delphatine prepared from lycoctonine are identical. Of interest is the fact that there is not a single, naturally occurring CI9-diterpenoid alkaloid which bears a C ( 1)-8-OCH3 group." (17) The configuration of the C(1)-methoxyl group must be also revised from @ to a in delbiterine (16-demethyldelphatine) and in elatine (7,8methy1encdioxy)methyllycaconitine.

Total Synthesis of (+)-Compactin (ML-236B)

PhS'

4, R = CHCH,OC,H, PhCH20

PhCH20

3, R = CHCH,OC,H,

5

Since chemical reduction of 6 with diisobutylaluminum hydride afforded a statistical mixture of racemic diastereomers, we decided to effect the reduction of 6 into 7 by using microbial methods. Although microbiological conversion of 6 to 7 had been reported,' the yield of this transformation was very low. On the other hand, exposure of 6 to Aureobasidium pullulans NRRL Y -1261O8 afforded 7, in 33% yield, accompanied by 8 (22%) and (f)-9 (45%). The optical purities of 7 and 8 were confirmed by their oxidation with Jones reagent to give (-)-6 and (+)-6, respectively. The inherent symmetry element (C2 axis) in this chiral intermediate, 7, and its ready accessibility via microbial methods markedly facilitated the ensuing transformations. Treatment of

I

Nai-Yi Wang, Chi-Tung Hsu, and Charles J. Sih* School of Pharmacy, University of Wisconsin Madison, Wisconsin 53706 Received July 6, 1981

Compactin' (1) and ML-236B are identical fungal metabolites isolated from strains of Penicillium brevicompactum and Penicillium citrinum, respectively. Compactin has been shown to be a potent competitive inhibitor2 of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA reductase), the rate-controlling enzyme in cholesterol biosynthesis.

1,R=H 2, R = CH, (1) Brown, A. G.; Smale, T. C.; King, T. J.; Hasenkamp, R.; Thompson, 1976, 1165. (2) Endo, A.; Kuroda, M.; Tanzawa, K. FEES Lett. 1976, 72, 323.

R. H. J . Chem. Soc., Perkin Tram. 1

0002-7863/81/1503-6538$01.25/0

..

0

..

RO

7,R=H 10, R = CH,Ph

8, R, = OH; R, = H 9, R, = H; R, = OH

7 with 2 equiv of NaH in MezSO at 25 OC, followed by the addition of C6H&H2C19 gave 10 (99%) as an oil, which upon reaction with phenylselenenyl bromide (HOAc/KOAc, 25 OC, 1 h) furnished 11. Saponification of 11 [1.2 equiv of KOH/ CH30H-ether (4:1), 25 OC, 2 h] afforded 12 (89%), which was oxidized to the corresponding selenoxide (H202/THF,25 "C, 2.5 h). The latter undergoes smooth elimination on heating at 55 O C for 2 h to give the allylic alcohol 13, which without isolation was subjected to Jones oxidation to give 5 (71% from 10). (3) Eaton, P. E.; Cooper, G.F.; Johnson, R. C.; Mueller, R. H. J . Urg. Chem. 1972, 37, 1947. (4) Posner, G. H.; Brunelle, D. J.; Sinoway, L. Synthesis 1974, 662. (5) Crossley, N. S. Tetrahedron Lett. 1971, 3327. (6) Van Tamelen, E. E.; Shamma, M.; Burgstahler, A. W.; Wolinsky, J.; Tamm, R.;Aldrich, P. E. J . Am. Chem. Soc. 1969, 91, 7315. Ireland, R. E.; Marshall, J. A. J. Org. Chem. 1962, 27, 1620. (7) Aklin, W.; Prelog, V.; Schenker, F.;Serdarevic, B.; Walter, P. Helu. Chim. Acta 1965, 48, 1725. (8) We thank Dr. C. P. Kurtzman of Northern Regional Research Center in Peoria for the identification of this soil isolate (Bu 103). (9) Iwashige, T.; Saeki, H. Chem. Pharm. Bull. 1967, 15, 1803.

0 1981 American Chemical Society

J. Am. Chem. Soc., Vol, 103, NO. 21, 1981 6539

Communications to the Editor PhCH20

PhCH20

PhCHz0

PhCH20

product.14 Hence, 21 was treated with an excess of MBA yielding 23 (97%). When the latter was hydrolyzed (20-fold excess KOH/EtOH, 25 "C, 16 h), 24 (63%) and 22 (6.6%) were obtained, accompanied by 23 (6.5%) and a trace of diol 21. Mesylation (MsCl/pyridine, 0 "C, 8 h) of 24, followed by elimination (DBU/pyridine, 115 "C, 3 h) furnished 25 (86%). The protecting group was removed by hydrolysis (HOAc/THF/H20, 3:2:1) to yield 26, which upon oxidation (PCC/CH2C12, 25 "C, 2 h) gave the aldehyde 27 (60%).

*OH

11, R = OAc 12, R = OH

13

When 5 was reacted with 4 (THF, -20 "C, 30 min), followed by the addition of excess methyl iodide (DME," -20 "C, 15 min), 14 (80%) was obtained. To ascertain the stereochemistry of the newly introduced methyl substituent, the conjugate addition was repeated but the intermediary copper-lithium enolate was alkylated with a slight excess of gaseous formaldehyde (THF, -78 "C, 30 min) to give a mixture (67%) of 15 and 16. This mixture was mesylated (MsC1/Et3N/CH2Cl2,-5 "C, 30 min) and the crude mesylate was subjected to elimination (DBU/benzene, 25 "C) to yield 17 (95% from 14). When 17 was hydrogenated (10% Pd/C, H1, EtOH), two epimers 3 and 14 were formed (75% yield) in a 5545 ratio, accompanied by 20% of 18. These were separated by repeated silica gel column chromatography (benzene/EtOAc, 93:7). In the presence of sodium methoxide in methanol (24 h, 25 "C), both epimers equilibrated to give a 9:l mixture of 14 and 3. Because 17 has a strong tendency to undergo isomerization on the surface of palladium metal, the hydrogenation of 17 was later conducted in the presence of pyridine" to minimize this undesirable side reaction. To our delight, under this condition the desired axial epimer 3 was obtained in 81% yield; concomitantly, formation of the equatorial epimer 14 was significantly diminished (10%).

14, R, = H; R , = C H , ; R , = 0 15, R , = H; R, = CH,OH; R, = 0 16, R, = CH,OH; R, = H; R, = 0 17. R , = R, s CH,; R , = 0 19, R , = Cf19;R, Hi R, = N-N(H)-Ph-CH,

18

, .

Condensation of the ketone 3 with tosyl hydrazidet2(anhydrous EtOH, 25 "C, 7 h) afforded 19, which after removal of the solvent was treated with 10 equiv of LDA (THF, -78-0 " C ) to yield 20 (62% from 3). Debenzylation (Li/liquid NH3) of 20 went smoothly giving 21 in quantitative yield. Attempts to selectively monoacylate 21 using 1 equiv of (fl-2-methylbutyric anhydrideI3 (MBA) (pyridine/DMAP, 25 "C, 16 h) gave 22 as the major (10) Coates, R. M.; Sandefur, L. 0. J . Org. Chem. 1974, 39, 275. (11) Djerassi, C.; Romo,J.; Rosenkranz, G. J. Org. Chem. 1951, 16, 754. (12) Commercial tosyl hydrazide must be pretreated with Et3N to neutralize a trace of acid, thus preventing the hydrolysis of the protecting group in 19. ( 1 3) (S)-2-Methylbutanol (Aldrich), -5.8' (neat) was oxidized (Von E. Doering, W.; Aschner, T. C. J . Am. Chem. SOC.1953, 75, 393) to (S)-2-methylbutyric acid, [a]*'D +18O (neat), which was converted (DCC/ CH2C12,0 OC, 3 h) to (S)-2-methylbutyric anhydride, bp 60 O C (0.2 mm); [.I2'D +30.2" (neat, d = 0.9318) in 97% yield.

H I

f0Y0V

20, R, = R, = CH,Ph 21,R,=R,=H 22, R , = H; R, = C(=O)CH(CH,)C,H, 23, R,C(=O)CH(CH,)C,H, = R, = 24, R, = C(=O)CH(CH,)C,H,; R, = H

25, R , =OCHCH,OC,H,; R, = H 26, R, = O H ; R , = H 27, R, = R , = O 28, R , = O H ; R , = CH,C(=O)CH,CO,CH, 29, R, = CH,C(=O)CH,CO,CH,; R, = O H

The lactone was incorporated into the bicyclic nucleus, 27, by reacting 27 with the dianion of methyl acetoacetate (NaHlnBuLi/THF, 0 "C, 15 min)" to yield two diastereomeric alcohols 28 and 29 (68%). Because this mixture resisted separation by TLC and HPLC, these two 6-hydroxy-@ketoesters were reduced [excess Z I I ( B H ~ )0~ "C, , 1 h] to two pairs of &G-dihydroxy esters (63%), 30 and 31 (3:2) [EtOAc/hexane (3:2)]. The less polar pair of the B,G-dihydroxy esters, 30 was separated from 31 by chromatographing the mixture over a silica gel column (acetone/CHC13, 1:9). Lactonization (p-TsOH-H20/benzene,25 "C, 30 min) of 30 afforded two /?-hydroxy lactones (65%) from which 1 (30 mg) was isolated by HPLC.I6 The synthetic specimen was found to be identical ([a]D,IR, HPLC, NMR, MS) with natural (+)-compactin. This synthetic sequence leading to compactin can also be adapted to the synthesis of mevinolin and other analogues, either directly or by chemical modification of intermediates at varying levels of functional and structural developments. These studies and further refinements of this scheme are currently under investigation.

Acknowledgment. This research was supported in part by a grant from the National Institutes of Health (HL 25772). We thank Dr. A. G. Brown of Beecham for a sample of natural compactin and Dennis Ditullio for conducting the microbiological work. Supplementary Material Available: TLC, mp, IR, 90-MHz 'H and I3C NMR, and [ a ] D data of new compounds (4 pages). Ordering information is given on any current masthead page. (14) The assignment of the position of the ester was based on the observation that mesylation and subsequent elimination of 22 gave a diene system whose NMR pattern was completely different from that of compactin. (15) Huckin, S.N.; Weiler, L. J. Am. Chem. SOC.1974, 96, 1082. (16) HPLC separation was carried out on a Waters radial compression module (RCM-100) with a radial-Pak 5 p silica gel cartridge (0.8 X 10 cm) (mobile phase, CHC1,; flow rate, 5 mL/min). The retention times of compactin and the second diastereomer in 30 were 16.4 and 21.8 min, respectively.